Gig workers, who are also considered as the genies of the COVID-19 influenced essential economy, have long been abandoned in the lending ecosystem due to their temporary, freelance jobs wanting steady paychecks. However, as their numbers have significantly grown in the world’s largest economies, lenders are now curious to expand their access to credit. Against the backdrop, artificial intelligence (AI) holds promise to convert gig workers into creditworthy borrowers by putting to use alternative data, says GlobalData, a leading data and analytics company.

An analysis of GlobalData’s Fintech Landscape Analytics database, which tracks fintech providers in the banking, payments, and insurance industries, uncovers various startups using AI-based alternative credit scoring to determine the creditworthiness of a thin file. Singapore’s Lenddo crunches multiple alternative data points, including e-commerce transactions, financial transactions, telecom, browser, mobile, social and psychometric data, for credit scoring of an applicant. Another Singapore startup CredoLab uses opt-in smartphone metadata to predict delinquent behavioral patterns of borrowers, which are then fed into their credit scoring. UK’s Credit Kudos uses real-time behavioral data to predict an applicant’s credibility to pay back a loan. It analyzes their interactions with digital devices such as mouse movements, keystroke pressure, gyroscopic movement and touch screen usage to scan for any potential fraudulent behavior. San Francisco-based

While banks are watching the opportunity slip through their hands, fintech startups are trying hard to exploit the gig economy that is rapidly growing with innumerable lending needs to be addressed

fintech startup Qwil, which tracks payment data in the gig economy, works with online marketplaces and service providers to pay freelancers on their behalf where they can have early access to funds rather than when their invoices are due. Another US-based startup Upstart evaluates educational data such as scholastic assessment test (SAT) score and grade point average (GPA) to predict the creditability of a loan seeker.

“The traditional credit scoring models, such as FICO, are inherently flawed in accessing thin credit files due to their assessment of only a handful of standard data variables. As a result, lenders often reel under pressure to make more inclusive credit decisions in real time. This is where fintech startups have come into action with their AI credit scoring models almost instantly interpreting alternative data like historical payments, digital footprint and behavioral economics,” says Kiran Raj, Principal Disruptive Tech Analyst, GlobalData.

“While banks are watching the opportunity slip through their hands, fintech startups are trying hard to exploit the gig economy that is rapidly growing with innumerable lending needs to be addressed. Of course, data privacy concerns creeping around alternative data can’t be ignored but can be sorted with evolving regulations in many countries. Moreover, gig workers are increasingly willing to share their data for inclusive credit decision making. Before it is too late, banks should start exercising AI-based alternative credit scoring models to expand their access to credit as well as to give a boost to financial inclusion,” concludes Kiran.